LGNO: A Local-Global Neural Operator for Hyperbolic Conservation Laws
Abstract
Solutions of hyperbolic conservation laws exhibit both smooth structures across large scales and sharp localized features such as shocks and contact discontinuities, making them difficult to approximate accurately with existing neural operators. The Fourier Neural Operator (FNO) captures long-range interactions well but tends to smear localized structures through excessive numerical dissipation. To address this, we propose a Local-Global Neural Operator (LGNO) that learns a one-step discrete flow map by combining a global FNO branch for representing smooth dynamics at large scales with a local multiresolution branch for enhancing localized discontinuities and nonsmooth features. The model is trained with a one-step loss that combines a physical space prediction term and a spectral penalty on high frequencies to suppress spurious oscillations near steep fronts. On a large collection of benchmarks in one and two dimensions, LGNO consistently outperforms FNO baselines with matched parameter counts, reducing one-step errors by factors of 2-5 and remaining significantly more accurate over long autoregressive rollouts. Most strikingly, although it is trained only on short-time data from a high-order WENO-Z scheme, the long-time rollout of LGNO on a coarse 2562 grid exhibits lower numerical dissipation than the same WENO-Z scheme run on a finer 5122 grid, while being orders of magnitude cheaper to evaluate. These results suggest that, with an appropriate architecture and training objective, learned operators can effectively learn discrete flow maps. They further suggest that such learned operators have the potential to control long-time numerical dissipation better than the conventional shock-capturing schemes that generate the training data.
Related
- A fully convergent fixed-point fast sweeping method with the WENO-JS local solver for steady state of hyperbolic conservation laws
- Local and global conservation laws of partial differential equations on moving domains in three dimensions
- Local gradient neural operator
- CLNO: Conservative Local Neural Operator for Compressible Flow Prediction
- Overcompressive shocks in a 2×2 hyperbolic system of conservation laws
BibTeX
@misc{wang2026lgno,
title = {{LGNO: A Local-Global Neural Operator for Hyperbolic Conservation Laws}},
author = {Hao Wang and Chi-Wang Shu and Qi Tang},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2606.18221},
eprint = {2606.18221},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2606.18221},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - LGNO: A Local-Global Neural Operator for Hyperbolic Conservation Laws
AU - Hao Wang
AU - Chi-Wang Shu
AU - Qi Tang
PY - 2026
DA - 2026/10/15
PB - arXiv
DO - 10.48550/arxiv.2606.18221
UR - https://arxiv.org/abs/2606.18221
AB - Solutions of hyperbolic conservation laws exhibit both smooth structures across large scales and sharp localized features such as shocks and contact discontinuities, making them difficult to approximate accurately with existing neural operators. The Fourier Neural Operator (FNO) captures long-range interactions well but tends to smear localized structures through excessive numerical dissipation. To address this, we propose a Local-Global Neural Operator (LGNO) that learns a one-step discrete flow map by combining a global FNO branch for representing smooth dynamics at large scales with a local multiresolution branch for enhancing localized discontinuities and nonsmooth features. The model is trained with a one-step loss that combines a physical space prediction term and a spectral penalty on high frequencies to suppress spurious oscillations near steep fronts. On a large collection of benchmarks in one and two dimensions, LGNO consistently outperforms FNO baselines with matched parameter counts, reducing one-step errors by factors of 2-5 and remaining significantly more accurate over long autoregressive rollouts. Most strikingly, although it is trained only on short-time data from a high-order WENO-Z scheme, the long-time rollout of LGNO on a coarse 256^2 grid exhibits lower numerical dissipation than the same WENO-Z scheme run on a finer 512^2 grid, while being orders of magnitude cheaper to evaluate. These results suggest that, with an appropriate architecture and training objective, learned operators can effectively learn discrete flow maps. They further suggest that such learned operators have the potential to control long-time numerical dissipation better than the conventional shock-capturing schemes that generate the training data.
ER -